Optimization of sampling basic unit size on spatial sampling for estimating winter wheat sown acreage

Sampling basic unit size is considered as a key factor in the process of formulating spatial sampling survey schemes. Reasonable sampling unit size plays an important role for reducing sampling survey cost and improving estimating accuracy of population value. The experiment on optimization of survey unit size was conducted to improve the efficiency of spatial sampling for estimating winter wheat sown acreage. Henan Province was selected as study area and Square grids as survey unit. The principle that it requests independent between sampling units in sampling frame is followed, and then the correlation is analyzed between survey unit sizes with Global spatial autocorrelation index (Moran I) in study area, in order to realize the primary selection of survey unit size. Based on the chosen survey unit sizes, 9 kinds of sampling frame are set up, and a stratified sampling scheme, which the stratification symbol is the ratio of winter wheat sown area accounting for the area of a survey unit, is used to choose samples, and then extrapolate population value as well as estimate sampling error. The experimental results demonstrate that, area sampling fraction increases with survey unit size increasing; the relative errors from population total estimate vary from 3.82% to 5.75% at 9 survey unit size levels; the spatial sampling efficiency is the highest using the survey unit size of 20000m × 20000m, where sampling survey cost and sampling error are selected as evaluation index of sampling efficiency.